RSV infection is the leading cause of infant hospitalization in industrialized countries. Following primary RSV infection, which generally occurs under the age of 2 years, immunity to RSV remains incomplete, and reinfection can occur. Furthermore, RSV can cause serious disease in the elderly and is, in general, associated with higher mortality than influenza A in non-pandemic years (Falsey et al., 1995). The WHO-estimated global annual infection rate in the human population is estimated at 64 million cases, with a mortality figure of 160000; in the US alone, from 85000 to 144000 infants are hospitalized each year as a consequence of RSV infection (on the World Wide Web at who.int/vaccine_research/diseases/ari/en/index2.html update 2009).
RSV belongs to the family Paramyxoviridae, subfamily Pneumovirinae, genus Pneumovirus; in human, there are two subgroups, A and B. Apart from the human RSV, there is a bovine variant. The genome of human RSV is approximately 15200 nucleotides long and is a negative-sense RNA molecule. The RSV genome encodes 11 known proteins: Glycoprotein (G), Fusion protein (F), Small hydrophobic protein (SH), Nucleoprotein (N), Phosphoprotein (P), Large protein (L), Matrix protein (M), M2 ORF-1 protein (M2-1), M2 ORF-2 protein (M2-2), Nonstructural protein 1 (NS1) and Nonstructural protein 2 (NS2). G, F and SH are transmembrane surface proteins; N, P, L, M, M2-1 are nucleocapsid associated proteins; and NS1 and NS2 are non-structural proteins. The status of M2-2 as a structural or nonstructural protein is unknown. (Hacking and Hull, 2002.) The RSV subgroups show differences in the antigenic properties of the G, F, N and P proteins (Ogra, 2004).
RSV infection is followed by the formation of specific IgG and IgA antibodies detectable in the serum and some other body fluids. Several studies have demonstrated that antibody responses are mainly directed to the major RSV transmembrane proteins F and G; only F- and G-specific antibodies are known to have in vitro RSV-neutralizing activity. Antibody responses to the F protein are often cross-reactive between the A and B subgroups, whereas antibody responses to the G protein are subgroup specific (Orga, 2004). Contrary to F and G, the transmembrane protein SH is considered as non-immunogenic (Gimenez et al., 1987; Tsutsumi et al., 1989) and in some vaccine candidates, SH has even been deleted in order to obtain a non-revertible attenuated vaccine (Karron et al., 2005).
Development of vaccines to prevent RSV infection has been complicated by the fact that host immune responses appear to play a significant role in the pathogenesis of the disease. Early attempts at vaccinating children with formalin-inactivated RSV showed that vaccinated children experienced a more severe disease on subsequent exposure to the virus as compared to the unvaccinated controls (Kapikian et al., 1969). Live attenuated vaccines have been tested, but show often over- or underattenuation in clinical studies (Murata, 2009).
Subunit vaccines using one immunogenic protein or a combination of immunogenic proteins are considered safer, because they are unable to revert or mutate to a virulent virus. Candidate vaccines based on purified F protein have been developed and were tested in rodents, cotton rats, and humans, and were shown to be safe, but only moderately immunogenic (Falsey and Walsh, 1996; Falsey and Walsh, 1997; Groothuis et al., 1998). In a similar vein, clinical trials with a mixture of F-, G- and M-proteins have been discontinued in phase II (ADISinsight Clinical database). An alternative approach consisted of a recombinant genetic fusion of the antigenic domain of human RSV G protein to the C-terminal end of the albumin-binding domain of the streptococcal G protein (BBG2Na; Power et al., 2001). BBG2Na was investigated up to a phase III clinical trial in healthy volunteers, but the trial had to be stopped due to the appearance of unexpected type 3 hypersensitivity side effects (purpura) in some immunized volunteers (Meyer et al., 2008).
A recent development is the use of chimeric recombinant viruses as vector for RSV antigens. A chimeric recombinant bovine/human parainfluenzavirus type 3 (rB/HPIV-3) was engineered by substituting in a BPIV-3 genome the F and HN genes by the homologous genes from HPIBV-3. The resulting chimeric rB/HPIV-3 strain was then used to express the HRSV F and G genes (Schmidt et al., 2002). This vaccine is currently under clinical investigation.
Only a limited number of prevention and treatment options are available for the severe disease caused by RSV. The most widely used intervention is based on passive immunoprophylaxis with a humanized monoclonal antibody that is derived from mouse monoclonal antibody 1129 (Beeler and van Wyke Coelingh, 1989). This antibody is specific for RSV F protein and neutralizes subgroup A and B viruses. The recombinant humanized antibody 1129 is known as palivizumab (also known as Synagis) and is used for prophylactic therapy of infants that are at high risk of developing complications upon RSV infection. The antibody is administered intramuscularly on a monthly basis in order to lower the risk of RSV infection in infants at risk due to prematurity, chronic lung disease, or hemodynamically significant congenital heart disease (Bocchini et al., 2009). Some studies have reported acceptable cost-effectiveness ratios for RSV prophylaxis with palivizumab (Prescott et al., 2010).